digi.ai.in
← digi.ai.in / blog
2026-09-237 min readPerformance

Why We Design for the Second Visit, Not Just the First

A well-cached second visit is faster, cheaper to win, and more likely to convert than any first visit — here's the evidence.

Most website briefs are written around the first visit: the landing page, the hero image, the pitch a stranger sees in the first five seconds. We think that's the wrong visit to optimise for. The second visit — when a browser already has your assets cached, when a user already half-trusts you — is faster to load, cheaper to win, and converts at a much higher rate than the first. This post lays out why, with numbers.

Repeat visits are technically a different product

A second visit isn't just an emotional repeat. Under the hood, it's a different technical event. Google's own web performance documentation is direct about this: "back and forward navigations, by definition, are repeat visits, and repeat page loads are generally faster than page loads from first time visitors due to HTTP caching," and Chrome's back/forward cache (bfcache) can restore a page instantly instead of reloading it — Google notes that "the page loads being replaced by bfcache restores would likely have been some of the fastest page loads in your dataset" (web.dev).

If a site adds a service worker on top of ordinary caching, the gap widens further. In Google's own case study, "visits to their app when controlled by a service worker loaded quite a bit faster than non-controlled visits, even those from returning users who likely had most of the page's resources cached" (web.dev). That's not a small optimisation — it means a returning visitor on a well-built site gets rewarded twice: once by the browser's default caching, and again by any extra caching layer the developer has deliberately engineered.

Speed also has a second-order effect on discovery. Since July 2018, Google has used page speed as a mobile search ranking factor (Google Search Central). A site engineered for fast repeat loads through caching isn't just pleasant to return to — it is more likely to rank well in the first place, because the same performance signals feed both experiences.

An Indian case study that proves the point

The clearest evidence isn't hypothetical. Treebo, the Bengaluru-based hotel brand, rebuilt its site as a Progressive Web App using service worker caching so JavaScript assets could be reused on repeat visits. The documented results, from Google engineer Addy Osmani's write-up:

Metric Improvement
Time to first paint 70%+ faster
Time-to-interactive 31% faster
Load time on 3G Under 4 seconds for many typical visitors
Interactive time (WebPageTest 3G emulation) Under 5 seconds

This matters specifically for India because the gains were measured against Indian 3G network conditions, not idealised lab conditions. If your users are opening your site on patchy 4G in a smaller city, this is the kind of gap that caching-for-return-visits closes.

The retention math is not close

Setting performance aside, there's a straightforward commercial reason to design for the second visit: it is dramatically cheaper to keep a visitor than to win a new one. Estimates on the exact multiplier vary, which is worth being honest about:

  • One industry analysis states "acquiring a new customer is consistently found to be 5 to 25 times more expensive than retaining an existing one," and cites Bain & Company research that "increasing customer retention rates by a mere 5% can boost profits by 25% to 95%" (Yotpo).
  • A separate source puts the figure more simply: "the cost of acquiring new customers is five times higher than the cost of retaining existing customers" (Optimove).

The two sources disagree on whether it's a flat 5x or a range up to 25x. Treat this as an order-of-magnitude signal rather than a precise number — but even the conservative end of that range makes the case that a site designed to bring people back is doing cheaper, more durable work than one designed only to convert a stranger once.

Returning visitors convert differently, not just more often

The gap shows up directly in conversion data, not just theory. A benchmark study across 24 direct-to-consumer stores found that "returning visitors are 18.8% of DTC sessions but 36.5% of revenue, converting 2.60x better than new" visitors (Top Growth Marketing). That's a small share of traffic generating twice its proportional share of revenue.

Separately, industry data suggests "a healthy eCommerce store typically maintains around 30% returning visitors and 70% new visitors," with "grocery and consumables seeing much higher repeat rates (up to 65%)" (Opensend). The same source notes that "baseline conversion rates for new visitors range from 1.5–2.5% across eCommerce" — a low ceiling for what a first-visit-only strategy can realistically achieve.

Headline ecommerce conversion figures vary by source. Shopify cites Statista data that "1.4% of global ecommerce visits converted into purchases in Q1 2026" (Shopify), while the DTC benchmark panel above implies a blended rate closer to 1.17%. Neither number should be read as universal — conversion rate is measured differently depending on traffic mix, device split, and industry. But both point the same direction: a single first visit is a weak place to expect a sale.

Why this matters more, not less, in India

India's internet base is enormous and still growing quickly. TRAI's FY2023–24 annual report shows "total number of internet subscribers increased from 88.1 crore at the end of March 2023 to 95.4 crore at the end of March 2024 with yearly rate of growth of 8.30%" (PIB). By Q2 FY2025, TRAI recorded "971.50 million internet subscribers... Out of the total 971.50 million internet users, 43.64 million were wired internet subscribers, while the majority, 927.86 million, were wireless internet subscribers" (source). Kantar's ICUBE 2024 survey, cited by IAMAI, found rural India now has more active internet users than urban India, out of 886 million active users nationally (IAMAI/Kantar report).

That user base is overwhelmingly mobile and still mostly on 4G. Nokia India's report found "mobile data consumption reached a record 24.1 GB per month for an average user in 2023," up from 19.5 GB in 2022, and "the 4G network contributed 85.2 percent to the total traffic, while the 5G network's share was 14.8 percent" (Forbes India). Data is cheap — at India Mobile Congress 2025 the Prime Minister noted "the cost of one GB of wireless data is less than the price of a cup of tea" (TelecomTalk) — but cheap data doesn't remove network variability, especially outside major metros. This is exactly the condition Treebo's caching strategy was tested against.

What this means in practice

If you're commissioning a website or app, here's what "designed for the second visit" should look like in a technical brief, in order of impact:

  1. Cache static assets aggressively. Images, fonts, CSS and JS that don't change per-user should be served with long cache lifetimes so a returning browser skips re-downloading them (web.dev).
  2. Consider a service worker for app-like experiences. Especially for anything with repeat transactional use — bookings, orders, dashboards — a service worker layer can meaningfully outperform standard caching on repeat visits (web.dev).
  3. Test on throttled 3G/4G, not just office wifi. Treebo's numbers were measured under Indian network conditions specifically because that's where the gap between a cached and uncached repeat visit is largest (Treebo case study).
  4. Track returning-visitor conversion separately from overall conversion. If returning visitors aren't converting at a meaningfully higher rate than new ones, something in the return experience — speed, personalisation, saved state — is being left on the table (Top Growth Marketing).
  5. Budget retention work like acquisition work. If retaining a visitor costs a fraction of acquiring one, even under the most conservative estimate, a design budget that spends everything on first-visit polish and nothing on repeat-visit performance is misallocated (Optimove).

None of this replaces a good first impression. But a first impression that isn't followed by a fast, familiar second visit is a wasted acquisition cost.

Questions people also ask

Does page speed actually affect Google ranking, or is that outdated advice?

It's current and specific: Google confirmed page speed became a mobile search ranking factor starting in July 2018 (Google Search Central). It's one signal among many, but it is real and has been in place for years.

Is it really 5 times cheaper to retain a customer than acquire one?

Sources vary — one puts it at a flat 5x, another says the range runs from 5x to 25x depending on the business (Yotpo, Optimove). Treat the number as directional rather than exact, but the direction is consistent across sources.

Does this matter more for mobile-first Indian users specifically?

Yes, for two reasons. Most of India's near-billion internet users are on wireless connections (TRAI data), and the majority of that mobile traffic still runs on 4G rather than 5G (Forbes India). Caching strategies that speed up repeat visits have measurably larger impact on those networks, as shown in Treebo's documented 3G test results (case study).

Written by the studio behind digi.ai.in. Questions? hi@digi.ai.in or WhatsApp +91 98605 96905
WhatsApp us